You still should consider collateral damage - we've been throwing broad spectrum antibiotics at damned near everything for decades. We've started an arms race

You still should consider collateral damage - we've been throwing broad spectrum antibiotics at damned near everything for decades. We've started an arms race with a staggeringly dynamic enemy, and our solution thus far has been to attack them with static munitions, until those munitions stop working. You can mitigate the issue with combination antibiotic therapy (it's harder to evolve resistance on multiple fronts simultaneously), but in general (when we use them) we need to move towards targeted antibiotics. The difficulty being that you actually need to diagnose the specific problem before treatment, not just that it's bacterial in nature. This requires pathogen detection and recognition: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4941824/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4941824/ This adds significant complexity (though I don't think that knowing what you're doing is a bad idea). We've been using nukes for a long time now, we need to move to scalpels.